Improvement of Power Quality in a Three-Phase System Using an Adaline-Based Multilevel Inverter

Frontiers in Energy Research

Prakash K. Ray¹  Arun Kumar Sahoo²  Soumya Ranjan Das²  Gaddam Sanjeeva Reddy³  Karthik Balasubramanian⁴ 

[1] Department of Electrical Engineering, College of Engineering and Technology, Bhubaneswar, India;Department of Electrical Engineering, International Institute of Information Technology Bhubaneswar, Bhubaneswar, India;Department of Electrical and Computer Engineering, Bhuma Shoba Nagireddy Memorial College of Engineering and Technology, Kurnool, India;Offshore Technology Development Private Ltd., Keppel Offshore and Marine Ltd., Singapore, Singapore;

关键词: Adaline-based LMS;    multi-level inverter;    recursive least square;    total harmonic distortion;    power quality;   

DOI  :  10.3389/fenrg.2020.00023

来源: DOAJ

【 摘 要 】

The existence of harmonics in the power distribution system (PDS) is treated as the most serious issue that affects its stability and reliability. The active power filter (APF) therefore plays a vital role in PDS to compensate for the harmonics for the improvement of the power quality (PQ) and to keep the total harmonic distortion (THD) below 5% as per IEEE-519. In this work, a three-phase four-wire (3P-4W) multi-level inverter (MLI)-based APF is proposed to overcome the shoot-through effect (STE) and reduce the distortions in the supply current. The control of the voltage source inverter (VSI) is achieved using an Adaline-based LMS (A-LMS) algorithm with a hysteresis current controller. The proposed filter that uses the A-LMS technique is compared with the conventional recursive least square (RLS) algorithm. The A-LMS approach is mainly operated for maintaining the dc link voltage of MLI, which follows the principle of capacitor energy and reduces the total harmonic distortion (THD) under different load variations. The performance of MLI under different load conditions is designed, developed, and validated by using a MATLAB/Simulink environment, and the preeminent features are established.

【 授权许可】

Unknown